An expert system for comprehensive diagnosis in store management system

Aiming at the condition that the fault mechanism of a store management system is complicated and it is difficult for diagnosis, an expert system for the equipment based on both neural networks and generate rules is discussed. At first, expert experience including fault phenomenons and reasons are summarized. Considering that expert knowledge is miscellaneous and is contact with too many parts, a new coding strategy is proposed by adopting Classification processing methods. NN is the primary tool of establishing Inference model and generate rules are mainly used for assistance, fault diagnosis algorithm of the expert system is designed. Then the expert system is realized based on Matlab Guide tool, results of the simulation proved the methods effective.